Eliciting broadly protective nAbs (bnAbs) against HIV-1 remains an enormous challenge for vaccine developers. Without exception, all vaccine candidates described to date have failed to elicit bnAbs against relevant tier 2 isolates similar t those that transmit infection. Fundamental problems could underlie this failure: it is a poorly documented fact that nearly all vaccine candidates tested to date fail to elicit detectable tier 2 nAbs against the autologous strain, let alone any tier 2 breadth. Like vaccine-induced Abs, the vast majority of Abs that develop during natural HIV-1 infection also fail to neutralize tier 2 strains. However, unlike vaccinated subjects, autologous nAbs do eventually develop during the course of most HIV-1-infections and bnAbs later evolve in ~25% of cases. Although this process is slow and inconsistent, it is the only known paradigm of bnAb development. Consequently, the evolutionary steps of nAb breadth development have been studied intensively. We are pleased to report that our trimer VLP vaccine candidates can predictably induce potent autologous (JR-FL) tier 2 nAbs. This result validates our earlier hypothesis that nAbs might best be elicited by the same antigen (viz. native trimer) that they must bind to neutralize the virus. However, we have yet to induce broadly cross- neutralizing tier 2 nAbs (bnAbs). We hypothesize that bnAbs might be induced by boosts designed to modify autologous nAb specificity in a manner that mimics key events in the evolution of breadth during natural infection. Our Specific Aims are: Specific Aim 1: To characterize a panel of glycan-modified VLP priming immunogens. A panel of trimer VLPs that grant variable access to known bnAbs but no access to non-nAbs will be evaluated for use as priming immunogens in Aims 3 and 4. Specific Aim 2: To improve VLP vaccine production methods. To address increasing demands for high VLP doses and a more sophisticated process, VLP manufacture will be modified to: i) maximize Env trimer expression, ii) allow high throughput concentration via filtration and iii) improve safety. Specific Aim 3: To improve vaccine-elicited nAb responses in rabbits. We seek to improve nAb consistency, potency and breadth in rabbits. Prime-boost immunizations will be designed to mimic aspects of bnAb development in natural infection, where viruses bearing slightly underglycosylated trimers (primes) induce autologous nAbs and are later boosted by fully glycosylated trimers and later by heterologous trimers. Thus, we hope to identify a pathway that incrementally evolves nAb breadth. Preliminary data supports the feasibility of this approach. Specific Aim 4: To improve vaccine-elicited nAbs in NHPs. Strategies from Aim 3 will be adapted to try to improve nAb responses in NHPs. NHP acquisition, immunizations and bleeds will all be conducted at the VRC at no cost. We request support only for VLP production and serum analysis.